IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
2879
Quaternary InGaAsSb Thermophotovoltaic Diodes
Michael W. Dashiell, John F. Beausang, Hassan Ehsani, G. J. Nichols, David M. Depoy, Lee R. Danielson,
Phil Talamo, Kevin D. Rahner, Edward J. Brown, Steven R. Burger, Patrick M. Fourspring, William F. Topper, Jr.,
P. F. Baldasaro, Christine A. Wang, Robin K. Huang, Member, IEEE, Michael K. Connors,
George W. Turner, Member, IEEE, Zane A. Shellenbarger, Gordon Taylor, Member, IEEE,
Jizhong Li, Ramon Martinelli, Member, IEEE, Dmitry Donetski, Member, IEEE,
Sergei Anikeev, Gregory L. Belenky, Fellow, IEEE, and Serge Luryi, Fellow, IEEE
Abstract—Inx Ga1−x Asy Sb1−y thermophotovoltaic (TPV)
diodes were grown lattice matched to GaSb substrates by metal–
organic vapor phase epitaxy in the bandgap range of EG = 0.5
to 0.6 eV. InGaAsSb TPV diodes, utilizing front-surface spectral
control filters, are measured with thermal-to-electric conversion
efficiency and power density (PD) of ηTPV = 19.7% and
PD = 0.58 W/cm2 , respectively, for a radiator temperature of
Tradiator = 950 ◦ C, diode temperature of Tdiode = 27 ◦ C,
and diode bandgap of EG = 0.53 eV. Practical limits to TPV
energy conversion efficiency are established using measured
recombination coefficients and optical properties of front surface
spectral control filters which for 0.53-eV InGaAsSb TPV energy
conversion are ηTPV = 28% and PD = 0.85 W/cm2 at the above
operating temperatures. The most severe performance limits are
imposed by 1) diode open-circuit voltage (VOC ) limits due to
intrinsic Auger recombination and 2) parasitic photon absorption
in the inactive regions of the module. Experimentally, the diode
VOC is 15% below the practical limit imposed by intrinsic Auger
recombination processes. Analysis of InGaAsSb diode electrical
performance versus diode architecture indicates that VOC and
thus efficiency are limited by extrinsic recombination processes
such as through bulk defects.
diode in order to recuperate below bandgap radiation [5], [6].
Quaternary InGaAsSb alloys were investigated because they
can be grown lattice matched to GaSb substrates for bandgaps
as low as 0.5 eV [7]–[11]; to date, however, the material
underperforms compared to ternary InGaAs TPV diodes. This
paper summarizes the theory used to predict the practical TPV
thermal-to-electric energy conversion efficiency for heat transferred radiatively from a hot-side radiator to a cold-side diode
module. Our analysis uses measured minority carrier recombination coefficients to determine the intrinsic limits to 0.53-eV
InGaAsSb TPV diode power conversion efficiency and assesses the electronic material quality and architecture (Table I)
required to approach these bounds. Fig. 1 shows the architecture
of a typical InGaAsSb n-p junction diode investigated during
this paper. A TPV module refers to the combination of front
surface spectral control filter and the TPV diode or an array
of diodes.
Index Terms—Diodes, indium gallium arsenide antimonide,
photovoltaic.
II. TPV EFFICIENCY (ηTPV )
I. INTRODUCTION
L
OW-BANDGAP (low-EG ) thermophotovoltaic (TPV)
converters have attracted interest in the field of direct
energy conversion due to the potential for efficient electric generation [1]–[3]. The best reported TPV efficiencies, measured
at Tradiator = 950 ◦ C and Tdiode = 27 ◦ C, are ηTPV = 24%
for EG = 0.6 eV InGaAs/InP [4] and ηTPV = 19.7% for
EG = 0.53 eV InGaAsSb/GaSb diodes [3]. Both systems utilized spectral control filters mounted on the front surface of the
Manuscript received May 16, 2006; revised August 26, 2006. The review of
this paper was arranged by Editor P. Panayotatos.
M. W. Dashiell, J. F. Beausang, H. Ehsani, G. J. Nichols, D. M. Depoy,
L. R. Danielson, P. Talamo, K. D. Rahner, E. J. Brown, S. R. Burger,
P. M. Fourspring, W. F. Topper, Jr., and P. F. Baldasaro are with Lockheed
Martin Corporation, Schenectady, NY 12301-1072 USA.
C. A. Wang, R. K. Huang, M. K. Connors, and G. W. Turner are with
Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, MA
02420-9108 USA.
Z. A. Shellenbarger, G. Taylor, and R. Martinelli are with Sarnoff Corporation, Princeton, NJ 08543-5300 USA.
J. Li is with the Center for Optical Technologies, Lehigh University, Bethlehem, PA 18015 USA.
D. Donetski, S. Anikeev, G. L. Belenky, and S. Luryi are with the State
University of New York, Stony Brook, NY 11794-2350 USA.
Color versions of Figs. 1–5 are available online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TED.2006.885087
The TPV thermal-to-electric power conversion efficiency
ηTPV (Table II) is defined as the electrical power output from
the TPV module divided by the total thermal power absorbed
in the module. The thermal power transferred from the hot
(Tradiator ) radiator to the cold module (Tdiode ) is due to the
radiative heat transfer across a vacuum gap. The maximum
electrical power from the TPV diode is the product of the
open-circuit voltage (VOC ), the short circuit current (ISC ),
and the fill factor (FF) [12]. Because photons having energies
(E < EG ) cannot be converted into electricity, it is convenient
to evaluate the overall efficiency as the product of the diode
efficiency ηDiode and the spectral efficiency ηSpectral [1], [3].
The term ηDiode quantifies the power conversion efficiency for
the above bandgap (E > EG ) thermal radiation absorbed in the
diode’s active n-p junction area. The term ηSpectral quantifies
the ratio of the above bandgap radiation absorbed in the active
area divided by the total thermal radiation absorbed in the
module. The breakdown of these efficiency parameters is given
in Table III. The overall expression for TPV thermal-to-electric
power conversion efficiency ηTPV is
ηTPV = ηSpectral × ηDiode
0018-9383/$20.00 © 2006 IEEE
=
ATotal
∞
0
VOC × ISC × FF
2πE
εcavity
dE
eff
h3 c2 (eE/kTradiator −1)
3
(1)
2880
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
TABLE I
SUMMARY OF 0.53-eV InGaAsSb ARCHITECTURES AND MEASURED OPEN-CIRCUIT VOLTAGES. ∗ REPORTED INTERFACE RECOMBINATION VELOCITIES
ARE AN AVERAGE VALUE OF FRONT AND BACK INTERFACES IN THE DH LIFETIME STRUCTURES IN REFS. [15]–[17]
values of InGaAsSb material coefficients [15]–[17] and the
optical properties of front surface spectral control filters [5],
[6] to determine a practical limit to TPV conversion efficiency
applied to the 0.53-eV InGaAsSb TPV material system.
III. MODELING ASSUMPTIONS
This section, along with Tables III and IV, summarizes
the TPV diode and spectral modeling assumptions used in
this paper.
A. Spectral Control Modeling Assumptions
Fig. 1. Schematic of typical 0.53-eV n-p InGaAsSb TPV diode architectures
grown on p-type GaSb substrates. p-n diodes having thick p-type emitters were
grown on n-type GaSb substrates.
where E, h, k, c, and Tradiator are the photon energy, Planck’s
constant, Boltzmann’s constant, the speed of light, and the
radiator (hot side) temperature, respectively. The effective cavmodifies Planck’s spectral distribution
ity emissivity εcavity
eff
to account for nonblackbody emission and absorption in the
optical cavity.
While thermodynamic analysis [1], [13], [14] can provide
a theoretical maximum to TPV conversion efficiency, nonradiative electronic recombination in the diode and parasitic
absorption in the module must be quantified both with experimental material data and mathematical models. This paper uses
established semiconductor theory and empirically determined
Due to optical absorption losses, particularly for photon
energies below the diode bandgap, the best reported spectral
efficiencies are ηSpectral ∼80% for the 0.53-eV bandgap and
Tradiator = 950 ◦ C [5], [6]. For simplicity, spectral performance calculations in this analysis assume a step function
reflection profile where the filter reflects 97% below bandgap
photons, reflects 15% above bandgap photons, and has a ∼2%
parasitic absorbance of above bandgap photons in the filter at all
incident photon angles. These values project a practical limit to
0.53-eV TPV spectral efficiency based on the optical properties
and design optimization studies of front surface filters. The
practical limit to spectral efficiency using the approximations
detailed in Table III is determined to be ηSpectral = 87%.
B. TPV Diode Recombination Models
The net radiative recombination rate per unit volume, calculated from the Shockley van Roosbroeke (SvR) detailed
balance method [18], must be corrected because some of the
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
2881
TABLE II
MEASURED THERMAL-TO-ELECTRIC EFFICIENCIES FOR 0.53-eV InGaAsSb TPV DIODE MODULES
TABLE III
SPECTRAL ASSUMPTIONS MADE FOR EG = 0.53 eV TPV PREDICTIONS
light emitted during radiative recombination will be reabsorbed
(recycled) in the active region of the TPV diode. The net
radiative recombination rate used in the simulations is given by
RRad =
B
np − n2i
ϕ
(2)
where Bn2i is the net thermal equilibrium rate of radiative
recombination per unit volume calculated via the SvR relation.
The photon recycling factor (ϕ) [19], [20] is the inverse ratio
of the sum of the photon flux exiting the diode’s front and back
surfaces to the total number of radiative recombination events
occurring within the diode volume (Table IV). The quantities n
and p in (2) are the electron and hole carrier densities under
illumination, respectively, and ni is the intrinsic carrier density.
Nonradiative recombination is a parasitic loss that includes
Auger recombination and recombination via bulk defect and
interface states that lie near the center of EG (Shockley-ReadHall (SRH) recombination). The net Auger recombination rate
is given by
RAug = (Cn n + Cp p) np − n2i
(3)
2882
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
TABLE IV
0.53-eV InGaAsSb DIODE PARAMETERS USED IN SIMULATIONS
and the bulk SRH and surface/interface recombination rates due
to electronic defect states near the center of the band gap (RSRH
and RSRV ) are modeled by
np − n2i
τn (p + ni ) + τp (n + ni )
RSRH =
(4)
RSRV =
Sn Sp np − n2i
.
Sn (p + ni ) + Sp (n + ni )
(5)
The recombination coefficients τn,p (SRH lifetimes), Sn,p (effective surface recombination velocities), and Cn,p (Auger coefficients) were parametrically varied in the simulations about
values reported in minority carrier lifetime studies of InGaAsSb
materials [15]–[17].
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
IV. INTRINSIC LIMITATIONS TO
TPV DIODE PERFORMANCE
This section discusses the intrinsic limits to TPV diode opencircuit voltage VOC , FF, and short circuit current density ISC .
A. Intrinsic Limitations to TPV Diode Open-Circuit
Voltage: VOC
The ideal diode equation expresses the open-circuit
voltage as
VOC
nkTdiode
ln
≈
q
Ilight − Io
Io
(6)
where Tdiode is the cold-side temperature, Ilight is the lightgenerated current, I0 is the dark current, q is the electron charge,
and n is the ideality factor.
The diode carrier generation/recombination rates that maintain thermal equilibrium with the cold side temperature Tdiode
determine the diode’s dark current (I0 )—also known as reverse
saturation current [1], [12]–[14]. Bounding the TPV diode
active region with a back surface reflector (BSR), rather than
an absorbing substrate or back contact, minimizes the radiative
component of I0 because BSR reduces the number of available photon modes that contribute to the overall equilibrium
generation/recombination rates [1] and sets the thermodynamic
maximum to VOC . In addition, the ideal BSR will increase the
optical path length by a factor of two, enabling a reduction
of the volumetric nonradiative recombination processes that
are unavoidable in real diodes. In concept, VOC limitations
due to additional extrinsic generation/recombination processes
such as bulk SRH defect recombination might be eliminated.
However, Auger recombination is a fundamental nonradiative
process that can be minimized, but not eliminated, by reducing
background carrier densities and absorber thickness. Attempts
to reduce Auger recombination limitations to VOC by reducing
background doping will ultimately result in the diode going into
high injection. The high injection Auger recombination rate
determines the lowest possible nonradiative recombination rate
under illuminated conditions and thus sets the intrinsic material
limit to I0 and VOC [21].
Using the narrow base approximations from [21], an estimate
for the values of the extrinsic coefficients Sn,p and τn,p required
to approach the intrinsic Auger-limited open-circuit voltage can
be made. Green [21] determined the VOC limits as a function of
light-generated current density JLight for absorbing region of
background doping NB and thickness W for each of the recombination processes given in (2)–(5). In a “defect-free” material,
the optimum background doping (NB ) and the intrinsic Augerlimited VOC are reached when the illuminated diode enters the
high-injection regime. For a p-type absorbing region, the bulk
electronic material quality is deemed sufficient to approach this
intrinsic limit only when the bulk SRH minority carrier lifetime
is greater than the quantity given by
1/3
2
(CP + Cn )
.
τphigh injection q 2 W 2 /JLight
(7)
2883
Anikeev et al. [15] report values of the Auger coefficient of
C ≈ 2 × 10−28 cm6 /s in 0.53-eV InGaAsSb. Thus, for lightgenerated currents of JLight ≈ 3 A/cm2 , the bulk SRH lifetime
must be greater than τn,p 1 µs in both n-type and p-type
InGaAsSb in order to approach the intrinsic Auger-limited
VOC . Using the same approach, the front and back surface/
interface recombination velocities need to be Sn,p ≤ 100 cm/s,
and the photon recycling factor must be ϕ 5 to approach the
intrinsic Auger-limited VOC . The last requirement means the
InGaAsSb TPV diode requires a BSR in order for the radiative
recombination rate to be small compared to the intrinsic Auger
recombination rate (see Table IV).
Measured minority carrier lifetimes in p-type double heterostructure (DH) InGaAsSb samples give bulk values of
τp ∼ 1 µs and Sp ≈ 700−2000 cm/s at the interfaces [15]–[17].
However, unlike p-type InGaAsSb, n-type InGaAsSb exhibits
a strong excitation dependence of the carrier lifetime that
shows the material is dominated by electrically active SRH-type
defects. The excitation dependence makes a precise measurement/model for the SRH lifetime at the relevant light injection
levels in n-type InGaAsSb (Te-doped) difficult; however, the
preliminary measurements set bounds of τn ≥ 0.1 µs in n-type
InGaAsSb and Sn ≤ 2000 cm/s at the n-type interfaces. Comparison of the measured n-type and p-type InGaAsSb recombination parameters with those values required to approach
the intrinsic Auger limit indicates that the electronic material
quality of 0.53-eV InGaAsSb is not sufficient to approach the
intrinsic Auger-limited VOC at these light-generated currents.
B. Intrinsic Limitations to TPV Diode Short Circuit
Current and FF
Unlike VOC , there are no thermodynamic restrictions to the
efficiency in which above bandgap photons can be converted to
photogenerated electrons. The primary limitations to Ilight ≈
ISC are due to parasitic absorption and reflections of above
bandgap photons in the spectral control filter and other inactive
regions of the TPV module (see Table III). Experimental data
for the 0.53-eV InGaAsSb diode indicate that the short circuit
current ISC is not limited by minority carrier recombination
in the active region because the minority carrier diffusion
lengths are greater than the diode thickness (narrow base),
and the effective surface recombination velocities are Sn,p <
104 cm/s. FF also depends on I0 ; however, unlike VOC , FF
becomes practically limited by series resistances external to
the diode active region. Experiments and calculations show that
2 mΩ · cm2 is the practical limit to series resistance setting an
FF limit of FF = 0.74 for JSC = 3 A/cm2 .
V. NUMERICAL SIMULATIONS OF 0.53-eV InGaAsSb TPV
EFFICIENCY AND OUTPUT POWER
Conversion of the above bandgap thermal radiation into
electric power was calculated using PC-1D, a numerical photovoltaic simulator that solves the carrier transport, Poisson, and
carrier continuity equations for electrons and holes. Solutions
were cross checked using the analytic expressions discussed in
this paper and in [12] and [21]. Nonequilibrium electron–hole
generation rates in the diode were determined from the
2884
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
Fig. 3. Simulated open-circuit voltage VOC for InGaAsSb TPV diode architectures as a function of doping using the recombination parameters listed in
Table IV. Each curve corresponds to the architecture described in the legend
and in the text.
Fig. 2. Intrinsic Auger-limited TPV conversion efficiency ηTPV and PD for
three different diode temperatures plotted as a function of Auger recombination
coefficient and using the spectral properties in Table III. Open markers illustrate
the limits for the case having nonnegligible extrinsic recombination losses.
absorbed thermal radiation spectrum transmitted through the
front surface spectral control filter into the diode.
Fig. 2(a) and (b) shows the intrinsic efficiency and power
density (PD) limits (solid symbols) as a function of the Auger
coefficients Cn,p for the 0.53-eV TPV diode module having
spectral control properties given in Table III, a BSR, negligible defect recombination, and operating at the onset to
high level injection. The measured Auger coefficient of Cp =
2 × 10−28 cm6 /s from [15] determines the best estimate to
the practical/intrinsic limit for 0.53-eV InGaAsSb TPV. Fig. 2
shows that by raising Tdiode from 27 ◦ C to 100 ◦ C, the practical efficiency limit drops from ηTPV = 28% to ηTPV = 18%
due to the strong temperature dependence of the intrinsic
carrier density. The open symbols in Fig. 2(a) and (b) show
the efficiency and PD limits for 0.53-eV TPV diodes having
nonnegligible extrinsic recombination losses (τn,p = 1 µs and
Sn,p = 500 cm/s), illustrating the stringent material requirements necessary to approach the intrinsic limit.
Fig. 3 shows the simulated diode VOC as a function of
p-type doping (NA ) for the general architecture shown in
Fig. 1. The uppermost curve in Fig. 3 corresponds to an
architecture having an ideal BSR and two-pass optical path
length (ϕ = 40) and negligible defect recombination (τn,p =
10 µs, Sn,p = 10 cm/s). The high-level-injection (NA ≤ 5 ×
1016 cm−3 ) Auger recombination rate sets the intrinsic limit to
VOC of 370 mV for the assumed operating temperatures. The
second uppermost curve simulates the same InGaAsSb material
quality (τn,p = 10 µs, Sn,p = 10 cm/s) but having an absorbing
substrate and single-pass optical path length (photon recycling
factor ϕ = 4), where the decrease in maximum achievable
VOC to 340 mV is due to the increased radiative dark current
and increased thickness to account for the single-pass optical
path length. The remaining curves show the degradation in
VOC for increasing extrinsic interface recombination losses
(Sn = Sp = S), demonstrating that the high-injection regime
is no longer optimal for diodes limited by defect recombination
and illustrating the severe VOC degradation. Similar trends
were observed when the TPV diode becomes bulk SRH defect
limited. The numerical simulations give good agreement with
the closed-form analysis in Section IV, stating that the extrinsic
parameters must be τn,p 1 µs and Sn,p ≤ 100 cm/s in order
to approach the intrinsic Auger limit.
VI. MEASURED 0.53-eV InGaAsSb TPV DIODE
ELECTRICAL/OPTICAL CHARACTERISTICS
Both n-p (thin-emitter) and p-n (thick-emitter) 0.53-eV
InGaAsSb DH TPV diodes were fabricated using standard
photolithography and metal evaporation [10], [22] having an
area of 0.5 cm2 . To prevent minority carriers from diffusing
to a free semiconductor surface, the diode is passivated with
either GaSb (EG ≈ 0.73 eV) or AlGaAsSb (EG ≈ 1.0 eV)
window layers [8]–[10], as shown in Fig. 1. Both GaSb and
AlGaAsSb back surface fields (BSFs) were investigated to
confine minority carriers at the rear of the active diode. A doped
GaSb contact layer was grown subsequent to the window layer
to enable ohmic contact formation for the metal contact grid.
The influence of parasitic series (rseries < 0.005 Ω · cm2 ) and
shunt resistances (rseries > 50 Ω · cm2 ) were negligible near
the operating conditions.
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
Fig. 4. Measured IQE for 0.53-eV n-p and p-n InGaAsSb diodes having a
thin GaSb window/contact layer and n-p diode having AlGaAsSb window and
thick GaSb contact layer. The inset shows a conceptual band diagram of the two
window interfaces.
A. Internal Quantum Efficiency (IQE) Measurements of
0.53-eV InGaAsSb TPV Diodes
Fig. 4 shows the measured (Tdiode = 27 ◦ C) IQE responses
for n-p and p-n 0.53-eV InGaAsSb TPV diodes. The IQE
response is the same for both n-p and p-n architectures having
thin (< 200 nm) GaSb window/contact layers (refer to figure
legend). The high IQEs between 1600 and 2250 nm indicate
a diffusion length for electrons (holes) in excess of the p-type
absorber thickness (thin n-type thickness), which is consistent
with simulations and [7] and [10]. The measured IQE degrades
at wavelengths below 1600 nm, corresponding to the GaSb
absorption cutoff wavelength [23]. Simulations indicate that
a significant fraction of minority carriers generated in the
GaSb contact layer may recombine at the unpassivated GaSbfree surface (Ssurface ≈ 105 cm/s) while the remainder may
diffuse into the active InGaAsSb region and be collected as
photocurrent. The experimental short-wavelength IQE response
degraded upon increasing GaSb contact thicknesses, which
supports this conclusion.
The third data set in Fig. 4 shows the measured IQE response
for an n-p architecture having a thick GaSb contact layer
(> 400 nm) and an AlGaAsSb passivating window layer. Severe
short-wavelength IQE degradation is observed in this third
curve for two reasons. First, the IQE experiences additional
degradation at short wavelengths because the GaSb contact
layer is more than twice as thick as for the previous two curves.
Second, the high EG of the AlGaAsSb layer relative to both
the GaSb contact layer and the InGaAsSb active region causes
it to act as a diffusion barrier to minority carriers generated
in the GaSb, which was confirmed by comparing measured
responses from two window materials having equivalent GaSb
contact thicknesses. The high EG AlGaAsSb window between
InGaAsSb active region and GaSb prevents nearly all minority
carriers generated in the GaSb contact from diffusing into the
active InGaAsSb, which degrades the short-wavelength IQE
response compared to using GaSb windows. The generic band
diagrams in the vicinity of the emitter for n-p diodes are shown
2885
Fig. 5. Plot of measured VOC versus JSC data (markers) for various light
illuminations for 0.53-eV p-n and n-p InGaAsSb TPV diodes. Inset provides
information on whether GaSb or AlGaAsSb was used for the passivating
window and BSF.
in the inset of Fig. 4, which illustrate these two mechanisms.
Note that the effect of this degradation on TPV efficiency and
short circuit current is small (<5% relative) due to the limited
radiation spectrum below 1600 nm expected from a 950 ◦ C
radiator.
B. Current–Voltage Measurements of n-p and p-n 0.53-eV
InGaAsSb TPV Diodes Having GaSb or InGaAsSb
DH Confinement
Fig. 5 shows the open-circuit voltage versus short-circuit
current density (VOC −JSC ) relation for six 0.53-eV InGaAsSb
TPV n-p and p-n diode architectures. The measured VOC
varies logarithmically with JSC , as expected from the ideal
diode model, and near-unity ideality (1.0 < n < 1.1) was
observed for all architectures for JSC values greater than
0.1 A/cm2 . The lowest measured dark currents were Jo =
1.5 × 10−5 A · cm−2 for p-n diodes having AlGaAsSb windows
and n-p diodes having either GaSb or AlGaAsSb windows.
For JSC > 0.1 A/cm2 , well below our expected operating currents, the current voltage characteristics for both n-p and p-n
architectures were insensitive to whether the BSF interface was
GaSb/InGaAsSb or AlGaAsSb/InGaAsSb. However, p-n diodes
with GaSb window layers have larger room temperature dark
currents of Jo = 2.5 × 10−5 A/cm−2 , which is attributed to the
greater than twofold increase in the front surface recombination
velocity. Measured surface recombination velocities at p-type
AlGaAsSb/InGaAsSb interfaces yielded average measured
values of Sp ∼ 700 cm/s, whereas the p-type InGaAsSb/GaSb
interfaces yield an average value of Sp ∼ 2000 cm/s [15]–[17].
The dependence of experimental J0 for diodes having p-GaSb
and p-AlGaAsSb minority confinement layers provides further
insight on the results in [15]–[17]. P-type AlGaAsSb front
surface windows reduce J0 in p-n diodes compared to p-type
GaSb windows; however, both p-type AlGaAsSb and GaSb
BSFs yield nearly equivalent J0 for n-p diodes. The asymmetry
indicates that the p-type AlGaAsSb passivation layer (due to its
higher EG ) suppresses the minority carrier diffusion to the free
2886
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
surface of the GaSb contact layer; however, its effect (if any) on
the number of defect states at the interface does not influence
J0 . Because AlGaAsSb causes a degradation in the IQE, an n-p
design using a thin GaSb window may be optimum for both
short wavelength photoresponse and high VOC .
Two simulated VOC −JSC curves are also shown in Fig. 5.
The first simulation (solid line) gives a reasonable agreement
(but slightly better) to the measured data by assuming recombination coefficients of Cn,p = 2 × 10−28 cm6 /s, τSRH = 1 µs,
Sn,p = 1000 cm/s, and an absorbing back surface (φ = 4). The
second simulation (dotted line) shows the practical limit for
a BSR diode architecture (φ = 40) having negligible extrinsic
recombination processes, illustrating the potential gains in VOC
with reduction in the extrinsic recombination losses.
C. Influence of Doping and Architecture on 0.53-eV InGaAsSb
TPV Diode Performance
A summary of the measured VOC (at JLight = JSC ∼
2.5 A/cm2 ) for various n-type and p-type doping levels and
minority carrier confinement layer compositions is shown in
Table I. Along with the carrier confinement layers are the corresponding estimates for the interface recombination velocities
Sn,p from [15]–[17]. The open-circuit voltage was normalized
to the bandgap (VOC /EG ) to account for slight variations
in alloy composition observed during the many growth runs
performed. Included also in Table I is the 0.53-eV InGaAsSb
hybrid BSR device reported in [24], which will have greater
photon recycling compared to that of a thick absorbing GaSb
substrate. All diodes listed in Table I had measured FFs of
0.7 ± 0.02 and peak IQEs near ∼100%.
The only significant change in VOC /EG was observed for
p-n architectures when the effective front surface recombination
velocity was increased from ∼700 to ∼2000 cm/s when using
the p-type GaSb window rather than AlGaAsSb. Table I shows
that VOC /EG remains unchanged for either n-type and p-type
doping ranging from mid 1016 cm−3 to low 1018 cm−3 . In
addition, VOC /EG also did not increase beyond a maximum
value of VOC /EG = 0.6 when the p-type AlGaAsSb/InGaAsSb
SRV was reduced to Sp ∼ 30 cm/s by optimizing the interfacial
growth conditions [17]. Finally, VOC did not increase upon
thinning the GaSb substrate and incorporating a BSR, although
a small increase in long wavelength quantum efficiency was
observed due to the increase in the effective optical path length
with the BSR [24].
The observed independence of the measured VOC /EG on
InGaAsSb doping levels ranging from mid 1016 cm−3 to low
1018 cm−3 (changing low injection Auger lifetime), optical
boundary conditions (changing photon recycling), and upon
reducing the surface recombination velocity to values near
Sp ∼ 30 cm/s, as well as the absolute value of the experimental
VOC , follows that behavior predicted for a diode whose VOC
is limited by extrinsic material recombination processes. The
experimental diode measurements are in agreement with the
discussion in Section IV, where the measured values of
the extrinsic recombination coefficients (τn,p and Sn ) do not
meet the requirements to obtain the practical limit to VOC .
Based upon this, we conclude that extrinsic defect recombina-
Fig. 6. Measured TPV diode dark current density versus EG for 0.5- to
0.6-eV InGaAsSb/GaSb and 0.6-eV InGaAs/InP. Experimentally determined
dark currents and standard deviations were obtained from batch-processed lots
of high-performance TPV diode.
tion mechanisms limit the InGaAsSb diode output voltage and
efficiency, in contrast to the conclusions in [7].
VII. 0.5- TO 0.6-eV InGaAsSb TPV DIODES LATTICE
MATCHED TO GaSb
Bandgap flexibility is an important consideration when
choosing a TPV diode material because the optimal diode
bandgap depends on both Tdiode and Tradiator . Fig. 6 shows
the experimental dark current J0 versus EG for p-n InGaAsSb
TPV diodes grown on GaSb substrates (circles). The ideality
factor for all InGaAsSb bandgaps was measured to be (1.0 <
n < 1.05), and all FFs were measured to be FF = 0.70 ± 0.02.
Measured InGaAsSb dark currents vary proportionally to
J0 α exp(−EG /kTdiode ), indicating that 1) the square of the
intrinsic carrier density dominates the InGaAsSb dark current
characteristics, and thus 2) the electronic material quality is not
significantly different over this bandgap range despite the material approaching the miscibility gap [9]. We note that the 0.6-eV
InGaAsSb diode probably has a slightly larger front SRV compared to lower bandgaps due to the smaller window band offset
[17]. As a qualitative materials comparison, an experimental J0
of 0.6-eV InGaAs/InP diodes, which is representative of 24%
efficient diodes [4], is shown in Fig. 6 (triangle). The dark
current for 0.6-eV InGaAs/InP lies well below that of the InGaAsSb/GaSb empirical data and fit, suggesting that the ternary
InGaAs possesses superior photovoltaic material properties
over the quaternary InGaAsSb alloy. High densities of anti-site
defects in antimonide-based semiconductors are responsible for
the observed high p-type background concentration, and donor
anti-site species have also been reported to be associated with
deep levels [25]. However, there is no sufficient published defect spectroscopy on low-EG InGaAsSb optoelectronic devices
to actually correlate anti-site defects with diode dark current.
Further work determining the defect structure in p-type and
n-type InGaAsSb and the SRH lifetime in n-type telluriumdoped InGaAsSb would provide additional data. Comparisons
of best SRH lifetimes (τp ∼ 1 µs) measured in InGaAsSb [15]
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
with values of (τp ∼ 5−14 µs) reported for GaAs photovoltaic
material [26], [27] are further evidence of the significant bulk
defect activity in the InGaAsSb alloy system.
2887
efficiency become more severe as the bandgap is reduced below
EG < 0.5 eV for the temperatures considered in this paper.
R EFERENCES
VIII. MEASUREMENTS OF 0.53-eV InGaAsSb
THERMAL-TO-ELECTRIC TPV CONVERSION
EFFICIENCY (ηTPV )
The performances of 0.53-eV InGaAsSb TPV diode modules
(1 and 4 cm2 areas) were measured in a prototypic vacuum test
cavity, as described in [3], [28], and [29]. The photonic cavity
is prototypical of a flat-plate TPV generator design, where the
radiator is a large flat silicon carbide surface. The TPV module
was fixed to the top of a copper pedestal to facilitate heat
absorption measurements. The thermal-to-electric efficiency in
this test was measured as the ratio of the peak module electric
power to the total module heat absorption rate. These parameters were measured simultaneously to assure validity of the final
efficiency value. A front surface filter, with spectral efficiency
calculated from reflection data to be ηSpectral ≈ 79%, is joined
to the modules with optical epoxy. Table II shows the measured
efficiency and electrical parameters of the 0.53-eV InGaAsSb
TPV module. As predicted, VOC , PD, and ηTPV decrease with
increasing temperature Tdiode . At 30 ◦ C, the average 0.53-eV
InGaAsSb TPV efficiency is only ∼70% of the practical
limit to efficiency based on the intrinsic Auger recombination
processes. The largest fractional difference in the experimental
performance versus the intrinsic limits is due to the diode
open-circuit voltage (VOC = 306 mV versus 370 mV), with the
remainder attributed to spectral performance (ηSpectral = 79%
versus 87%) and module FF (FF = 67% versus 74%) and the
percent active area (Aactive = 84% versus 90%). Based on the
previous discussion, extrinsic recombination processes limit
the InGaAsSb TPV diode VOC , making this the most significant
barrier to reaching the practical efficiency limits.
IX. CONCLUSION
A practical limit of ηTPV = 28% and PD = 0.85 W/cm2 for
0.53-eV InGaAsSb TPV diodes operating at Tradiator = 950 ◦ C
and Tdiode = 27 ◦ C is established. The most severe bound
to low-bandgap TPV diode performance will be set by the
intrinsic Auger limit to open-circuit voltage, which for 0.53-eV
max
= 370 mV. The measured InGaAsSb
InGaAsSb will be VOC
TPV efficiency and PD indicate that InGaAsSb TPV diodes
operate well below the intrinsic performance limits, primarily
because the diode VOC is dominated by extrinsic recombination processes such as through bulk defect levels. The semiempirical method used in this article can be applied to other
TPV systems and operating conditions, where Auger recombination becomes particularly important if the diode operating
temperature increases and/or the diode bandgap decreases,
since the Auger recombination coefficients and intrinsic carrier
density increase exponentially as EG is reduced and Tdiode is
increased. Using the ranges of Auger coefficients versus EG
reported in [7], our semi-empirical analysis determines that the
intrinsic Auger recombination limits to open-circuit voltage and
[1] P. F. Baldasaro, J. E. Raynolds, G. W. Charache, D. M. Depoy,
C. T. Ballinger, T. Donovan, and J. M. Borrego, “Thermodynamic analysis
of thermophotovoltaic efficiency and power density tradeoffs,” J. Appl.
Phys., vol. 89, no. 6, pp. 3319–3327, Mar. 2001.
[2] T. J. Coutts and J. S. Ward, “Thermophotovoltaic and photovoltaic conversion at high-flux densities,” IEEE Trans. Electron Devices, vol. 46, no. 10,
pp. 2145–2153, Oct. 1999.
[3] E. J. Brown, P. F. Baldasaro, S. R. Burger, L. R. Danielson, D. M. Depoy,
J. M. Dolatowski, P. M. Fourspring, G. J. Nichols, and W. F. Topper, “The
status of thermophotovoltaic energy conversion technology at Lockheed
Martin Corporation,” in Proc. Collect. Tech. Pap. Int. Int. Energy Convers.
Eng. Conf.. Washington, DC: Amer. Inst. Aeronautics and Astronautics,
2004, vol. 2, pp. 1296–1315.
[4] B. Wernsman, R. G. Mahorter, R. Siergiej, S. D. Link, R. J. Wehrer,
S. J. Belanger, P. Fourspring, S. Murray, F. Newman, D. Taylor, and
T. Rahmlow, “Advanced thermophotovoltaic devices for space nuclear
power systems,” in Proc. STAIF, 2005, vol. 746, pp. 1441–1448.
[5] T. D. Rahmlow, J. E. Lazo-Wasem, E. J. Gratrix, P. M. Fourspring, and
D. M. Depoy, “Design considerations and fabrication results for front surface TPV spectral control filters,” in Proc. 6th Conf. Thermophotovoltaic
Generation Elect., 2004, vol. 738, pp. 180–188.
[6] P. M. Fourspring, D. M. DePoy, T. D. Rahmlow, Jr., J. E. Lazo-Wasem,
and E. J. Gratrix, “Optical coatings for thermophotovoltaic spectral control,” presented at the Optical Interference Coatings, Tucson, AZ, 2004,
ThE10. CD-ROM (The Optical Society of America, Washington, DC).
[7] G. W. Charache, P. F. Baldasaro, L. R. Danielson, D. M. Depoy,
M. J. Freeman, C. A. Wang, H. K. Choi, D. Z. Garbuzov, R. U. Martinelli,
V. Khalfin, S. Saroop, and J. M. Borego, “InGaAsSb thermophotovoltaic
diode: Physics evaluation,” J. Appl. Phys., vol. 85, no. 4, pp. 2247–2252,
Feb. 1999.
[8] C. W. Hitchcock, R. J. Gutmann, J. M. Borego, I. B. Bhat, and
G. W. Carache, “Antimonide-based devices for thermophotovoltaic applications,” IEEE Trans. Electron Devices, vol. 46, no. 10, pp. 2154–2161,
Oct. 1999.
[9] C. A. Wang, H. K. Choi, S. L Ransom, G. W. Charache,
L. R.
Danielson, and D. M. Depoy, “High-quantum-efficiency
GaInAsSb/GaSb thermophotovoltaic devices,” Appl. Phys. Lett., vol. 75,
no. 9, pp. 1305–1307, Aug. 1999.
[10] H. K. Choi, C. A. Wang, G. W. Turner, M. J. Manfra, D. L. Spears,
G. W. Charache, L. R. Danielson, and D. M. Depoy, “High performance
GaInAsSb thermophotovoltaic devices with an AlGaAsSb window,” Appl.
Phys. Lett., vol. 71, no. 26, pp. 3758–3760, Dec. 1997.
[11] M. W. Dashiell, J. F. Beausang, G. Nichols, D. M. Depoy, L. R. Danielson,
H. Ehsani, K. D. Rahner, J. Azarkevich, P. Talamo, E. Brown, S. Burger,
P. Fourspring, W. Topper, P. F. Baldasaro, C. A. Wang, R. Huang,
M. Connors, G. Turner, Z. Shellenbarger, G. Taylor, J. Li, R. Martinelli,
D. Donetski, S. Anikeev, G. Belenky, S. Luryi, D. R. Taylor, and J. Hazel,
“0.52 eV quaternary InGaAsSb thermophotovoltaic diode technology,” in
Proc. 6th Conf. Thermophotovoltaic Generation Elect., 2004, vol. 738,
pp. 404–414.
[12] S. M. Sze, Physics of Semiconductor Devices, 2nd ed. Hoboken, NJ:
Wiley, 1981.
[13] W. Shockley and H. A. Queisser, “Detailed balance limit of efficiency
of p-n junction solar cells,” J. Appl. Phys., vol. 32, no. 3, pp. 510–519,
Mar. 1961.
[14] C. H. Henry, “Limiting efficiencies of ideal single and multiple energy
gap terrestrial solar cells,” J. Appl. Phys., vol. 51, no. 8, pp. 4494–4500,
Aug. 1980.
[15] S. Anikeev, D. Donetski, G. Belenki, S. Luryi, C. A. Wang,
J. M. Borrego, and G. Nichols, “Measurement of the Auger recombination
rate in p-type 0.54 eV GaInAsSb by time-resolved photoluminescence,”
Appl. Phys. Lett., vol. 83, no. 16, pp. 3317–3319, Oct. 2003.
[16] D. Donetski, S. Anikeev, G. Belenky, S. Luryi, C. A. Wang, and
G. Nichols, “Reduction of interfacial recombination in GaInAsSb/GaSb
double heterostructures,” Appl. Phys. Lett., vol. 81, no. 25, pp. 4769–
4771, Dec. 2002.
[17] D. Donetski, S. Anikeev, N. Gu, G. Belenky, S. Luryi, C. A. Wang,
D. A. Shiau, M. Dashiell, J. Beausang, and G. Nichols, “Analysis of recombination processes in 0.5–0.6 eV epitaxial GaInAsSb lattice matched
to GaSb,” in Proc. 6th Conf. Thermophotovoltaic Generation Elect., 2004,
vol. 738, pp. 320–328.
2888
[18] W. van Roosbroeke and W. Shockley, “Photon-radiative recombination of
electrons and holes in germanium,” Phys. Rev., vol. 94, no. 6, pp. 1558–
1560, Jun. 1954.
[19] P. Asbeck, “Self-absorption effects on the radiative lifetime in
GaAs–GaAlAs double heterostructures,” J. Appl. Phys., vol. 48, no. 2,
pp. 820–822, Feb. 1977.
[20] A. Marti, J. L. Balenzategui, and R. F. Reyna, “Photon recycling and
Schockley’s diode equation,” J. Appl. Phys., vol. 82, no. 8, pp. 4067–4075,
Oct. 1997.
[21] M. A. Green, “Limits on the open-circuit voltage and efficiency of silicon
solar cells imposed by intrinsic Auger processes,” IEEE Trans. Electron
Devices, vol. ED-31, no. 5, pp. 671–678, May 1984.
[22] R. Huang, C. A. Wang, C. T. Harris, M. K. Connors, and D. A. Shiau,
“Ohmic contacts to n-type GaSb and n-type GaInAsSb,” J. Electron.
Mater., vol. 33, no. 11, pp. 1406–1410, 2004.
[23] S. J. Adachi, “Optical dispersion relations for GaP, GaAs, GaSb, InP,
InAs, InSb, AlGaAs, and InGaAsP,” J. Appl. Phys., vol. 66, no. 12,
pp. 6030–6040, Dec. 1989.
[24] R. K. Huang, C. A. Huang, M. K. Connors, and G. W. Turner,
“Hybrid back surface reflector GaInAsSb thermophotovoltaic devices,”
in Proc. 6th Conf. Thermophotovoltaic Generation Elect., 2004, vol. 738,
pp. 329–336.
[25] A. G. Milnes and A. Y. Polyakov, “Review: Gallium antimonide device
related properties,” Solid State Electron., vol. 36, no. 6, pp. 803–818,
Jun. 1993.
[26] J. M. Olsen, R. K. Ahrenkiel, D. J. Dunlavy, B. Keyes, and A. E. Kibbler,
“Ultralow recombination velocity at GaInP/GaAs heterointerfaces,” Appl.
Phys. Lett., vol. 55, no. 12, pp. 1208–1210, Sep. 1989.
[27] L. W. Molenkamp, G. L. M. Kampschoer, W. deLange, J. W. F. M. Maes,
and P. J. Roksnoer, “Ultralong minority-carrier lifetimes in GaAs grown
by low-pressure organometallic vapor-phase epitaxy,” Appl. Phys. Lett.,
vol. 54, no. 20, pp. 1992–1994, May 1989.
[28] C. K. Gethers, C. T. Ballinger, and D. M. DePoy, “Thermophotovoltaic
efficiency testing,” in Proc. 4th Conf. Thermophotovoltaic Generation
Elect., 1998, vol. 460, pp. 335–348.
[29] B. Wernsman, R. R. Siergiej, S. D. Link, R. G. Mahorter,
M. N. Palmisiano, R. J. Wehrer, R. W. Shultz, R. L. Messham, S. Murray,
C. S. Murray, F. Newman, D. Taylor, D. Depoy, and T. Rahmlow, “Greater
than 20% radiant heat conversion efficiency of a thermophotovoltaic
radiator/module system using reflective spectral control,” IEEE Trans.
Electron Devices, vol. 51, no. 3, pp. 512–516, Mar. 2004.
[30] C. A. Wang, H. K. Choi, D. C. Oakley, and G. W. Charache, “Expressions
fit from mobility data of InGaAsSb layers grown on insulating GaAs
substrates given in C.A.,” J. Cryst. Growth, vol. 195, no. 1–4, p. 346,
1998.
Michael W. Dashiell was born in Baltimore, MD,
in 1970. He received the B.S., M.S., and Ph.D.
degrees from the University of Delaware, Newark,
in 1992, 1998, and 2001, respectively, all in electrical
engineering.
From 1992 to 1996, he was a Research Engineer
with AstroPower Inc. (currently GE Solar), where
he developed high-efficiency silicon and GaAs solar
photovoltaics. From 1997 to 2002, he was with the
Department of Electrical Engineering, University of
Delaware, and the Max-Planck-Institute for Solid
State Physics, Stuttgart, Germany, where he conducted research regarding
epitaxial growth, electrical and optical characterization of metastable silicon
and silicon-carbide-based alloys, tunnel diodes, and self-assembled quantum
dots. Since 2002, he has been an Electrical Engineer with the Knolls Atomic
Power Laboratory, Niskayuna, NY (operated by Lockheed Martin Corporation,
Schenectady, NY, for the U.S. Government), where he has been developing lowbandgap III–V thermophotovoltaic energy converters and high-performance
power electronic motor drives. He is also an Adjunct Professor with Union
Graduate College, Schenectady, and has authored or coauthored over 45 technical publications.
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
John F. Beausang was born in Hershey, PA, on
November 11, 1976. He received the B.S. degree in
engineering science from Pennsylvania State University, University Park, in 1999 and the M.S. degree in
physics from Rensselaer Polytechnic Institute, Troy,
NY, in 2002. He is currently working toward the
Ph.D. degree in physics at the University of Pennsylvania, Philadelphia.
Until 2004, he was with Lockheed Martin Corporation, Schenectady, NY, where he worked on
thermophotovoltaic technology.
Hassan Ehsani received the B.S. and M.S. degrees
in physics from the University of Tehran, Tehran,
Iran, in 1974 and 1976, respectively, the M.S. degree
from The University of Kansas, Lawrence, in 1984,
and the Ph.D. degree from The University of Texas,
Austin, both in electrical engineering, in 1988.
From 1976 to 1981, he was a faculty member with
the University of Tehran. From 1988 to 2000, he was
a Research Scientist with Rensselaer Polytechnic
Institute, Troy, NY. Since 2000, he has been a Senior Scientist with Knolls Atomic Power Laboratory,
Niskayuna, NY (operated by Lockheed Martin Corporation, Schenectady, NY,
for the U.S. Government). His research interests have covered a wide range of
semiconductor physics and optoelectronic materials including organometallic
chemical vapor deposition (OMCVD) growth of II–VI and III–V based compound semiconductors for infrared detectors, lasers, and thermophotovoltaics;
metallization of group II–VI and III–V semiconductors, thermophotovoltaics,
and solar cell devices; and processing, characterizations, and modeling. He is
the author or coauthor of over 80 technical publications.
G. J. Nichols, photograph and biography not available at the time of
publication.
David M. Depoy received the B.S. and M.S.
degrees from Rensselaer Polytechnic Institute, Troy,
NY, in 1989 and 1993, respectively, both in mechanical engineering.
Since 1989, he has been with the Advanced
Concepts Group, Knolls Atomic Power Laboratory, Niskayuna, NY (operated by Lockheed Martin
Corporation, Schenectady, NY, for the U.S. Government), investigating direct energy conversion
technologies. He has worked in the field of thermophotovoltaic (TPV) energy conversion for the past
13 years with emphasis on spectral control, characterization, performance modeling, and generator design. He is the author or coauthor of over 30 technical
publications and holds five U.S. patents.
Lee R. Danielson received the B.S. degree from
Iowa State University, Ames, in 1968, the M.S.
degree from the University of Washington, Seattle,
in 1969, and the Ph.D. degree from Washington State
University, in 1977, all in physics.
From 1979 to 1988, he was with Thermo Electron
Corporation, Waltham, MA, where he performed
research on higher efficiency thermionic and thermoelectric energy conversion. Since 1989, he has
been with Lockheed Martin Corporation, Schenectady, NY, where he has worked on thermoelectric and
thermophotovoltaic energy conversion and has developed several electrical and
thermal property measurement systems.
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
Phil Talamo received the B.S. degree in marine engineering systems from the United States Merchant
Marine Academy, Kings Point, NY, in 1984.
In 1984, he joined Knolls Atomic Power Laboratory, Niskayuna, NY (operated by Lockheed Martin
Corporation, Schenectady, NY, for the U.S. Government), where he is currently managing a group
conducting research and development of advanced
energy conversion technologies, specifically direct
energy conversion. The bulk of the research effort
is on thermophotovoltaics (TPV) with efforts to
improve the conversion efficiency of high-performance semiconductor TPV
devices and filters. His career started as an Operational Engineer at one of
the nation’s naval nuclear operating prototype, where naval personnel learn the
fundamentals of operation prior to being assigned to one of the nation’s nuclearpowered submarine or aircraft carrier. Prior to his current position, he managed
the training program at the nuclear prototype, responsible for the training of all
navy sailors.
Kevin D. Rahner received the B.S. degree in physics
from the University at Albany, State University of
New York (SUNY), Albany, and the M.S. degree in
electrical engineering from Binghamton University,
SUNY.
Since 2002, he has been an Engineer with Lockheed Martin Corporation, Schenectady, NY, working
on electrical performance modeling and radiative
heat transfer modeling of TPV systems.
Edward J. Brown received the B.S. degree in
physics and the M.S. degree in physics from the State
University of New York at New Paltz in 1979 and the
M.S. degree in nuclear engineering from Rensselaer
Polytechnic Institute, Troy, NY, in 1983.
In 1982, he joined the technical staff with Knolls
Atomic Power Laboratory, Niskayuna, NY (operated
by Lockheed Martin Corporation, Schenectady, NY,
for the U.S. Government), working in the areas of
reactor core design, reactor safety analysis, and experimental reactor operations. In 1990, he joined the
Advanced Concepts Group working on direct energy conversion. Since 1997,
he has supervised the energy conversion testing program supporting the development of thermophotovoltaic energy conversion technology. He holds four
U.S. patents. His areas of research interest include direct energy conversion,
nuclear reactor core design, and reactor dynamics modeling.
Mr. Brown is a member of Tau Beta Pi and Alpha Nu Sigma and was the
President of the NENY Section of the American Nuclear Society.
Steven R. Burger received the B.S. degree in chemical engineering from Manhattan College, Riverdale,
NY, in 1986 and the M.S. degree in mechanical
engineering from Union College, Schenectady, NY,
in 1986.
Since 1982, he has been with the Knolls Atomic
Power Laboratory, Niskayuna, NY (operated by
Lockheed Martin Corporation, Schenectady, for the
U.S. Government), where he is currently a Senior
Engineer in the Advanced Components and Systems
Unit working on turbomachinery design. He is the
author or coauthor of ten technical publications. His development and work
interests have covered a wide range of engineering disciplines including
semiconductor packaging development for direct energy conversion devices,
nondestructive test development for reactor plant components, reactor plant
fluid system design, and reactor plant equipment maintenance and modification.
2889
Patrick M. Fourspring received the B.S. and M.S.
degrees in structural mechanics from Clemson University, Clemson, SC, in 1982 and 1986, respectively,
and the Ph.D. degree in engineering science and
mechanics from the Pennsylvania State University,
University Park, in 1997.
His entire career has been associated with nuclear
power. From 1980 to 1990, he was with Duke Power
(now Duke Energy), where he was a Cooperative
Education Student supporting system modifications
for operating nuclear power stations and a Design
Engineer responsible for the identification and implementation of technology
to improve the performance of nuclear power stations. In 1994, he joined
the Knolls Atomic Power Laboratory, Niskayuna, NY (operated by Lockheed
Martin Corporation, Schenectady, NY, for the U.S. Government) and has
completed numerous mechanical analyses of the nuclear reactor components
as well as assessments of various nuclear reactor configurations. Since 2003,
he has been leading the development of spectral control technology for thermophotovoltaic energy conversion of heat from a nuclear reactor source.
Dr. Fourspring is a member of the American Society of Mechanical Engineers (ASME). He is a Registered Professional Engineer.
William F. Topper, Jr. received the B.S. degree
in mechanical engineering from Cornell University,
Ithaca, NY, in 1971, and the M.S. degree in mechanical engineering from Rensselaer Polytechnic
Institute, Troy, NY, in 1974. He is currently working
toward the Ph.D. degree at Rensselaer Polytechnic
Institute.
He has 35 years of professional experience covering developmental testing of gas turbine engine
components with AVCO Lycoming; program management for manufacturing, development, and design of nuclear system components with General Electric; and design and
professional engineering certification of medical waste incinerator and energy recovery systems with WFT Engineering Services. He is currently with
Lockheed Martin Corporation, Schenectady, NY, and has been in several
positions providing technical customer support, performing component design,
and working on development and testing of nuclear power plant equipment.
His current position involves the development of energy conversion systems
and components and the design and operation of developmental test apparatus
for advanced mass transport system components.
P. F. Baldasaro, photograph and biography not available at the time of
publication.
Christine A. Wang received the S.B., M.S., and
Ph.D. degrees from the Materials Science and Engineering Department, Massachusetts Institute of
Technology, Lexington, in 1977, 1978, and 1984,
respectively.
In 1984, she joined Lincoln Laboratory, Massachusetts Institute of Technology, Cambridge, where
she is currently a Senior Staff Member in the ElectroOptical Materials and Devices Group and works
on III–V materials grown by organometallic vapor
phase epitaxy (OMVPE) for advanced optoelectronic
devices. She is the author or coauthor of over 140 technical publications and
holds four U.S. patents. Her research interests include gas transport and design
of OMVPE reactors, growth and materials characterization of diode lasers,
thermophotovoltaic devices, and detectors.
Dr. Wang is a member of Tau Beta Pi, Sigma Xi, and the Materials Research
Society. She is currently the Vice-President of the American Association of
Crystal Growth.
2890
IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 53, NO. 12, DECEMBER 2006
Robin K. Huang (M’03) received the S.B. degree in
physics with electrical engineering from the Massachusetts Institute of Technology (MIT), Cambridge,
in 1995, and the M.S. degree in electrical engineering and the Ph.D. degree in applied physics from
Stanford University, Stanford, CA, in 1997 and 2000,
respectively. In 2000, he joined Lincoln Laboratory,
MIT, where he is currently a Staff Member. He is
the author or coauthor of over 35 technical publications. His research interests include high-power
near-infrared semiconductor diode lasers, thermophotovoltaic devices, mid-IR diode lasers, and vertical cavity surface emitting lasers (VCSELs).
Michael K. Connors received the B.S. degree in
biological science and environmental science from
the University of Massachusetts, Lowell, in 1977.
In 1977, he joined Lincoln Laboratory, Massachusetts Institute of Technology, Cambridge, where
he is currently an Assistant Staff Member in the
Electro-Optical Materials and Devices Group. He is
the coauthor of over 45 publications. His research
interests include development of device fabrication
processes, reactive ion etching of semiconductor
materials, and thin-film deposition of dielectric and
metal coatings.
George W. Turner (M’00) received the B.S., M.S.,
and Ph.D. degrees in electrical engineering from
Johns Hopkins University, Baltimore, MD, in 1972,
1975, and 1979, respectively.
In 1979, he joined Lincoln Laboratory, Massachusetts Institute of Technology, Cambridge, where
he has been a Staff Member in a number of different
groups in the Solid-State Division. He is currently an
Assistant Leader in the Electro-Optical Materials and
Devices Group. He authored or coauthored several
book chapters, over 125 technical publications, and
holds four U.S. patents. His research interests have covered a wide range of
device fabrication and material growth areas, including avalanche photodiodes,
direct energy conversion devices, and mid-infrared lasers.
Zane A. Shellenbarger received the M.S.E.E. degree from the University of
Delaware, Newark, and the B.A. degree in physics from Cornell University,
Ithaca, NY.
He was with AstroPower Inc. and Structured Materials Inc. for 11 years,
working on the development of epitaxial growth and the fabrication of thermophotovoltaics, solar cells, infrared detectors, and LEDs. In 2000, he joined
Sarnoff Corporation, Princeton, NJ, where he is currently a Member of Technical Staff and has been responsible for the epitaxial growth of InP-, GaAs-,
and GaSb-based lasers, thermophotovoltaic cells, and other optoelectronic
devices. He is the author of over 40 technical publications in journals and
conference proceedings on laser diodes, photovoltaics, thermophotovoltaics,
infrared detectors, and other optoelectronic devices. His technical expertise is
in the epitaxial growth of semiconductor materials by metalorganic chemical
vapor deposition and the characterization of semiconductor materials.
Gordon Taylor (S’76–M’76) received the B.S., M.S., and Ph.D. degrees in
electrical engineering (solid state device specialization) from Rutgers University, Camden, NJ, in 1969, 1972, and 1976, respectively.
He has been with Sarnoff Corporation, Princeton, NJ, for over 30 years.
Jizhong Li received the M.S. degree in optoelectronics and information engineering from Shandong University, Qingdao, China, in 1989 and the Ph.D.
degree in solid-state physics from Kansas State University, Manhattan, in 1999.
Before 1994, he was an Associate Professor with the Department of Physics,
Qufu Normal University, Qufu, China, where he conducted research on laser
beam modulation and polarization technologies. During his postdoctoral research with North Carolina State University in 2000, he conducted AlGaN/GaN
material growths by metal organic chemical vapor deposition (MOCVD) for
the fabrication of solar blind (250–280 nm) detector arrays. From 2000 to
2005, he was a member of Technical Staff with Sarnoff Corporation, where
he conducted the development of various III–V compound materials for fabricating high power lasers and thermophotovoltaic devices. He is currently a
Research Scientist with the Center for Optical Technologies, Lehigh University,
Bethlehem, PA.
Ramon Martinelli (S’59–M’84) received the A.B. and M.S. degrees in electrical engineering from Dartmouth College, Hanover, NH, in 1960 and 1961,
respectively, and the Ph.D. degree in electrical engineering from Princeton
University, Princeton, NJ, in 1965.
He is a Distinguished Member of Technical Staff with Sarnoff Corporation,
Princeton, NJ.
Dmitry Donetski (SM’98–M’97) received the
Diploma (with Honors) and the Ph.D. degree in
physics from St. Petersburg State Technical University, St. Petersburg, Russia, in 1987 and 1996, respectively, and the Ph.D. degree in electrical engineering
from the State University of New York, Stony Brook,
in 2000.
From 1987 to 1996, he was with the Department
of Semiconductor Physics and Nanoelectronics, St.
Petersburg State Technical University, working on
the design of far-infrared semiconductor lasers and
electrooptic modulators on hot carriers. In 1996, he joined the Optoelectronics
Group, Electronics and Communications Engineering Department, State University of New York, Stony Brook. His research is focused on the design and
technology of optoelectronic devices including thermophotovoltaic cells, midinfrared laser diodes, and laser diode arrays. He has 50 publications in referred
journals and conference proceedings.
Sergei Anikeev received the M.Sc. degree (with
the MSU Graduation medal) in physics from
Lomonosov Moscow State University (MSU),
Moscow, Russia, in 2000, and the Ph.D. degree in
electrical engineering from State University of New
York, Stony Brook, in 2006.
In 1998, he was awarded the scholarship of
Teaching-Research Center “Fundamental Optics and
Spectroscopy.” He is currently with the State University of New York. He authored or coauthored
15 papers in scientific journals and conference proceedings. His research interests include experimental and analytical study of
energy relaxation processes in semiconductors and physics of thermophotovoltaic devices.
DASHIELL et al.: QUATERNARY InGaAsSb THERMOPHOTOVOLTAIC DIODES
Gregory L. Belenky (M’95–SM’96–F’05) received
the Doctor of Physics and Mathematics degree in
Russia.
In 1991, he was with AT&T Bell Labs,
Murray Hill, NJ. In 1995, he joined the State University of New York, Stony Brook, where he is currently
a Professor in the Department of Electrical and Computer Engineering. He has published over 140 papers
and four reviews and filed four U.S. patents dealing
with the physics of two-dimensional structures and
physics and design of photonic devices. His current
interests include physics of semiconductors and the design and working performance of semiconductor lasers and optoelectronic systems.
2891
Serge Luryi (M’81–SM’85–F’89) received the
Ph.D. degree in physics from the University of
Toronto, Ontario, Canada, in 1978. His doctoral thesis was on the theoretical studies of intermolecular
interactions in solid hydrogen.
In 1980, he was with Bell Laboratories, Murray
Hill, NJ, and became interested in the physics and
technology of semiconductor devices. In 1994, he
joined the State University of New York, Stony
Brook, where he currently chairs the Electrical and
Computer Engineering Department. In 1995, he organized an advanced research workshop on the “Future Trends in Microelectronic,” (FTM) which grew into a regular series. The fifth FTM workshop took
place in June 2006 on the island of Crete, Greece (complete information can
be found at www.ee.sunysb.edu/~serge/FTM.html). He is also the Founding
Director of the NY State Center for Advanced Sensor Technology (Sensor
CAT). He has published over 200 scientific papers and holds 43 U.S. patents.
Dr. Luryi was elected Fellow of the IEEE for contributions in the field of
heterojunction devices in 1989, received the Distinguished Member of Technical Staff award from Bell Laboratories in 1990, and was elected Fellow of the
American Physical Society in 1993 for contributions to the theory of electron
transport in low-dimensional systems and invention of novel electron devices.
In 2003, he was appointed to the rank of Distinguished Professor by the Board
of Trustees of the State University of New York. He served as the Editor of
IEEE TRANSACTIONS ON ELECTRON DEVICES from 1986 to 1990.